New hardware platforms for MiSTer?

[iamTeddyNL]

Greetings MiSTer community members. My name is Teddy, I am an electronic engineer and a (retro) gaming enthousiast.

I am highly impressed by the technology developed by this community, and I see great potential for real life adaptations. However, I get the idea that the hardware that is currently available is not ideal to achieve that. Being a dev kit, it is not designed for a ‘production’ (real life application) environment, some features are missing that may impact reliability, it is not available as a complete solution and has to be assembled and sourced separately, and the only enclosure that I’ve seen (from sheet metal?) could do the job but isn’t visually appealing and probably also not cost effective. I get the idea that this project would benefit from new hardware that would offer a compatible, complete and reliable solution to end users. I have been flirting with the idea to raise a kickstarter campaign to support that, so I created an account here to introduce myself and to gather feedback about this idea.

Right now I have two products in mind that would facilitate real life applications:

- A console, similar to a Super NT but without the cartridge slot of course. Customers would receive a complete and enclosed unit, no assembly required and ready to be configured right away. Target audience would be retro gamers and content creators.
- An arcade computing unit, similar to an automotive computer/industrial pc, intended to be retrofitted into an arcade cabinet where the processing hardware has failed – in the end, all of them will fail eventually. With industrial grade assembly processes and components, it would be designed for a production environment with high reliability and long lifetime.

I have supported a kickstarter campaign for home automation before as a supplier, which has turned out to be a commercial success, and this has given me a network of European suppliers that is keen to support kickstarter campaigns – no outsourcing to Asia. This allows a much greater control over quality and availability, not succumbing to the poor availability trap that a lot of kickstarter campaign fall for. This does impact pricing so it will not be low cost, but seeing that the current total cost of ownership is already fairly high for the current hardware and competing platforms like the Analogue stuff, it seems viable to come up with pricing that is reasonable and works for all parties involved.

Before I come up with business cases, contacting suppliers and preparing/raising the actual campaign, I have come to do a bit of market research first to learn if this initiative would be commercially viable and supported by this community. I would like to get in touch with the leaders/key users/experts in this community to discuss this, can you please point me to the right direction?

Many thanks,
Teddy

[andrezheng]

Great idea. BTW, where can you get the 5CSEBA6 at a reasonable price?

[iamTeddyNL]

Great idea. BTW, where can you get the 5CSEBA6 at a reasonable price?

Depends on the quantities you need. For single pieces, the standard catalog houses like Digikey/Mouser are your only bet, unless you want to venture into unauthorized channels and risk counterfeit parts. They are expensive but that is for the service that they offer of repacking into single pieces. For volume, wholesale distributors come into play. You might want to make use of your subcontractor's purchasing channel for better conditions.

[Hackshed_Carl]

So you want to take an open source project and commercialize it?

For me, the best thing about the MiSTer platform is that the DE-10 Nano is very reasonably priced not to mention being supported by a huge company.
All addons are very affordable and can be designed by anyone.

It sounds to me like you're wanting to re-design the entire platform to something more like the MiST platform
I don't expect you'll get much support from this community for that

....just my 2 pence worth

[alanswx]

There are a few mister projects based on the de10-nano that accomplish your goals. They are a bit or an extreme design because they have to plug into the dev board. Jammix, mistercade, and there is a console now too.

[Duffygag]

Not sure you will get a lot of traction here

[iamTeddyNL]

So you want to take an open source project and commercialize it?

For me, the best thing about the MiSTer platform is that the DE-10 Nano is very reasonably priced not to mention being supported by a huge company.
All addons are very affordable and can be designed by anyone.

It sounds to me like you're wanting to re-design the entire platform to something more like the MiST platform
I don't expect you'll get much support from this community for that

....just my 2 pence worth

Commercialization / making money is not my objective. I am interested in, and pitching the idea for, a MiSTer in a Nintendo classic mini / Super NT form factor, maybe something for arcade too (if not already existing, the feedback from alanswx is helpful). This would never work in an 'open source' approach, like sharing the schematics/gerbers/STEP files and go have one made for yourself. Whenever you have hardware manufactured, commercial aspects come into play to make it sustainable and that is why I am mentioning it. Hope that clarifies things.

[alanswx]

Commercialization / making money is not my objective. I am interested in, and pitching the idea for, a MiSTer in a Nintendo classic mini / Super NT form factor, maybe something for arcade too (if not already existing, the feedback from alanswx is helpful). This would never work in an 'open source' approach, like sharing the schematics/gerbers/STEP files and go have one made for yourself. Whenever you have hardware manufactured, commercial aspects come into play to make it sustainable and that is why I am mentioning it. Hope that clarifies things.

As long as it is compatible with MiSTer I don't see a huge problem. I think many of the add on projects aren't open hardware.

MiSTerCade: https://misteraddons.com/products/mistercade
Antonio has a Jamma adapter as well: https://www.antoniovillena.es/store/pro ... a-adapter/
Jammix is both a Jamma adaper an an ITX form factor: https://twitter.com/jammixo?lang=en
MultiSystem: viewtopic.php?t=3031

there are a bunch of open source cases, and computer designs.

There are some commercial cases, including misteraddon's metal one: https://misteraddons.com/products/copy- ... se-round-2

I am sure I am leaving out a dozen other neat commercial and open source projects.

To date the de10-nano development board has subsidized the FPGA part, so everything is an add on board to that..

[vanfanel]

@iamTeddyNL

I find your vision VERY attractive: we really need something like the MiSTer but in SuperNT-like for, ready to use, no modular desing, NO catidge slots (microSD is enough), simply a board where all those cores can be deployed for a ~150 to 200 price tag. If that's doable, count on me.

BUT I hope you are not trying to TIVOlize the cores. If you do something based on the MiSTer project, keep it open or don't do it.
What we DO NOT need is another Analogue.

Whatever they tell you around here, make a single-board MiSTer-compatible product, and it will sell like there's no tomorrow, you can be sure.

[zakk4223]

@iamTeddyNL

I find your vision VERY attractive: we really need something like the MiSTer but in SuperNT-like for, ready to use, no modular desing, NO catidge slots (microSD is enough), simply a board where all those cores can be deployed for a ~150 to 200 price tag. If that's doable, count on me.

Right, the rub is that's not doable. Even if you somehow managed to find someone to give you a 50% volume discount on the FPGA, it would still cost around $125 just for that component alone. That's assuming you can even find volume stock, supply is very constrained right now and you might have a surprising lead time ahead of you.

I don't think many people realize exactly how good of a deal the DE10-Nano actually is considering the components involved...

[iamTeddyNL]

@iamTeddyNL

I find your vision VERY attractive: we really need something like the MiSTer but in SuperNT-like for, ready to use, no modular desing, NO catidge slots (microSD is enough), simply a board where all those cores can be deployed for a ~150 to 200 price tag. If that's doable, count on me.

Right, the rub is that's not doable. Even if you somehow managed to find someone to give you a 50% volume discount on the FPGA, it would still cost around $125 just for that component alone. That's assuming you can even find volume stock, supply is very constrained right now and you might have a surprising lead time ahead of you.

I don't think many people realize exactly how good of a deal the DE10-Nano actually is considering the components involved...

I haven't put effort into quoting that FPGA yet, but I very much doubt this is an FPGA that goes for $125 in volume. I have been in the electronic components industry for years, I am a professional and I know what I am doing. Also very much aware of the current situation with semiconductors, but those days will pass. It probably already has passed if I go ahead with this idea once I have the pcb and enclosure developed and tested and the molds tooled up. Pretty confident I can beat the €325 ex VAT price of the ultimate MiSTer Pro - A Super NT goes for much less and also utilizes a Cyclone V. And yes, my intention is for this board to be completely compatible with the DE10, with all the peripherals like the RAM or the RTC already on the board. Not a single hardware or software restriction in place.

Please don't worry about constraints like pricing or availability. I am requesting feedback on the concept mentioned earlier (a MiSTer in a console solution) and the price range of 150-200 is indeed what I had in mind. Should be doable if the demand is there.

[zakk4223]

Right. The Super NT uses a different Cyclone V, the 5CEBA4F23C8N. It has no ARM core and it has roughly half the logic elements. The single unit price of that is ~$53.

The Mister uses the SE 5CSEBA6U23I7 which has a single unit price of ~$243

I'm sure people will be ecstatic if you can hit the $200 price point. Good luck.

[dshadoff]

I'm sure you could get a better price than $243 in volume, but whether you'd be able to get cost or availability to an acceptable point is going to be based on your target price, volume, and existing relationship with Altera/Intel. Good luck !

[djb_rh]

It’s a great idea. I’m guessing you can even pull it off *if* you’re willing to buy thousands of FPGAs that will then take you several years to sell.

There are enough cores that are rock solid that if you could sell complete working systems WITH GAMES INSTALLED I could see you selling into the thousands quickly. But since you can’t do that, I think it’s more in the hundreds per year. I could be wrong, and it somewhat depends on what you built and how cheap you could get it, but as long as there’s still some hacking going on to get the games on it, it’s just not a huge seller.


—Donnie

[DevilHunterWolf]

I think with all the current and future MiSTer redesign concepts out there, the standard home MiSTer is well taken care of. And with MiSTerCade and other JAMMA conversion products out there, arcade platforms are pretty well covered too. A MiSTer is already a pretty small device even in a full build. Other than just making it look neat on a shelf, there's not really any big benefits to yet another redesign. I understand getting the price down is also a goal of this and I'm sure people will welcome a cheaper MiSTer build. But unless it's a significant price drop, it's going to be hard to beat the basic DE10-Nano and SDRAM starting components for most value for the dollar.

Personally, I think the better thing to shoot for is a portable MiSTer compatible. That's about the only thing that's really missing and why people (myself one of them) are still looking at Analogue's Pocket handheld. I'd love to have the MiSTer as a handheld, even if it was something as big as a GameBoy DMG. And that at least would give room to debate about the price. "Sure you could spend $400 USD on a MiSTer, but wouldn't you rather spend it on a *portable* MiSTer that can still connect to a TV?" It would be a different crowd to target, could get some double dippers from those that already have a MiSTer, and it would be more unique among the rest. That's my opinion, anyway.

[akeley]

About the price, just wanted to mention that a MiSTer setup which will run everything costs ~250 USD at the moment (RAM+DE10+cheap USB hub). And if you go with 32MB (which still runs 98% of the library) you are getting close to 200 USD mark.

Just saying this in general, not connected directly to OP, because there is this inflated price misconception which many newcomers take for granted.

[atrac17]

Bubblegum & taffy.

[aberu]

I haven't put effort into quoting that FPGA yet, but I very much doubt this is an FPGA that goes for $125 in volume. I have been in the electronic components industry for years, I am a professional and I know what I am doing. Also very much aware of the current situation with semiconductors, but those days will pass. It probably already has passed if I go ahead with this idea once I have the pcb and enclosure developed and tested and the molds tooled up. Pretty confident I can beat the €325 ex VAT price of the ultimate MiSTer Pro - A Super NT goes for much less and also utilizes a Cyclone V. And yes, my intention is for this board to be completely compatible with the DE10, with all the peripherals like the RAM or the RTC already on the board. Not a single hardware or software restriction in place.

Please don't worry about constraints like pricing or availability. I am requesting feedback on the concept mentioned earlier (a MiSTer in a console solution) and the price range of 150-200 is indeed what I had in mind. Should be doable if the demand is there.

I look forward to hearing what you can get in terms of a quote. If it's at around the same price as the DE10-Nano itself then it probably won't be in that much demand, but ya never know. People are strangely excited to pay a heck of a lot more for a behemoth software emulation platform like the Polymega (which still seems like vaporware, but I digress).

[andrezheng]

Is it possible to add an TFT LCD, like 4.3 inch LCD to make Mister portable.............

[softtest9]

You can already buy a full MiSTer build with SDRAM, USB, a case and everything from a few sellers. While it may not be perfect, I think it's good enough for most users.

In my opinion, what would really set your product apart, would be a better FPGA chip. I know that FPGAzumSpass for example cancelled his Nintendo DS core due to the lack of logic elements on the DE10-nano. Nintendo 64 and beyond is likely also impossible on the DE10-nano.

[aberu]

You can already buy a full MiSTer build with SDRAM, USB, a case and everything from a few sellers. While it may not be perfect, I think it's good enough for most users.

In my opinion, what would really set your product apart, would be a better FPGA chip. I know that FPGAzumSpass for example cancelled his Nintendo DS core due to the lack of logic elements on the DE10-nano. Nintendo 64 and beyond is likely also impossible on the DE10-nano.

Unfortunately, in order to do that basically every single core will have to be ported all over again, the framework would have to be ported. It would entail a lot more work than designing a complex PCB and case. there aren't really any Altera chips that fit that bill right now that also work with Quartus Lite (free). So whoever makes a possible "successor" or upgraded MiSTer in the future, would likely have to look to Xilinx.

Hopefully this hobby just keeps growing in this way, and I'm sure sorg has kept his eye on developments in the dev board space for this kind of thing.

[Blitzwing]

You can already buy a full MiSTer build with SDRAM, USB, a case and everything from a few sellers. While it may not be perfect, I think it's good enough for most users.

In my opinion, what would really set your product apart, would be a better FPGA chip. I know that FPGAzumSpass for example cancelled his Nintendo DS core due to the lack of logic elements on the DE10-nano. Nintendo 64 and beyond is likely also impossible on the DE10-nano.

Unfortunately, in order to do that basically every single core will have to be ported all over again, the framework would have to be ported. It would entail a lot more work than designing a complex PCB and case. there aren't really any Altera chips that fit that bill right now that also work with Quartus Lite (free). So whoever makes a possible "successor" or upgraded MiSTer in the future, would likely have to look to Xilinx.

Hopefully this hobby just keeps growing in this way, and I'm sure sorg has kept his eye on developments in the dev board space for this kind of thing.

It's also worth mentioning that the only reason this version of the Cyclone 5 was cheaply (relatively) accessible, with all the other needed components is because it is still subsidised by Intel through its university program. Even without all the other factors you mentioned, how cost prohibitive a more powerful FPGA would be is a step too far I think right now... I think a more realistic, but unlikely target might be a version that clocks higher to help out Ao486.

[softtest9]

Is the chip alone subsidized though? I thought the DE10-nano was subsidized, and Analogue for example are not using this board.

[Duffygag]

No need to reinvent the Mister, besides everyone is doing a different version of the same thing.

Do a Mister handheld and you will have everyone's support

[Blitzwing]

Is the chip alone subsidized though? I thought the DE10-nano was subsidized, and Analogue for example are not using this board.

I think the whole device is subsidised, I was just typing hurriedly as normal

Analogue probably have a production scale that allows them to reduce cost, plus whatever they do in house.

I suppose if a company like Terasic wanted to start making their own gaming type board then they could also reduce costs for a beefier device, but that doesn't seem their wheelhouse. Point being I think for a hobbyist project like MiSTer we're just going have to wait for costs to reduce and for a large company to bring out a new subsidised Dev board.

Edit. That goes without saying that something like the Cyclone 5 fits all the relatively basic machines, the next step is the OG Xbox, PS2, GC, Dreamcast and later arcades and they are a whole different can of worms.

[aberu]

You can already buy a full MiSTer build with SDRAM, USB, a case and everything from a few sellers. While it may not be perfect, I think it's good enough for most users.

In my opinion, what would really set your product apart, would be a better FPGA chip. I know that FPGAzumSpass for example cancelled his Nintendo DS core due to the lack of logic elements on the DE10-nano. Nintendo 64 and beyond is likely also impossible on the DE10-nano.

Unfortunately, in order to do that basically every single core will have to be ported all over again, the framework would have to be ported. It would entail a lot more work than designing a complex PCB and case. there aren't really any Altera chips that fit that bill right now that also work with Quartus Lite (free). So whoever makes a possible "successor" or upgraded MiSTer in the future, would likely have to look to Xilinx.

Hopefully this hobby just keeps growing in this way, and I'm sure sorg has kept his eye on developments in the dev board space for this kind of thing.

It's also worth mentioning that the only reason this version of the Cyclone 5 was cheaply (relatively) accessible, with all the other needed components is because it is still subsidised by Intel through its university program. Even without all the other factors you mentioned, how cost prohibitive a more powerful FPGA would be is a step too far I think right now... I think a more realistic, but unlikely target might be a version that clocks higher to help out Ao486.

There are ones out there which are roughly twice as powerful as the de10-nano's Cyclone V (and have a lot more BRAM which is key) but cost like 50% more, but would require a subcarrier board to be designed as well. Someone is looking into it, but I doubt it will be quite as strong a "competitor" (for lack of a better word) as they think it will be, and it could be a year or two at minimum, IF it even works out. It's very very cost-prohibitive compared to the MiSTer then as a result.